Do we really need an AI answer? They're trash. Go ask if the 737 has a RAT, or which large aircraft don't.
It's pretty clear that BTV isn't going to be safer than using max autobrake/max manual every time, in much the same way that using TOGA thrust is going to give the best safety on takeoff (barring some exceptions around Vmc), or taking off with full fuel tanks even if not necessary.
However, acceptable safety can be achieved with a system that brings a bunch of maintenance, comfort, and/or throughput advantages.
The 'obvious' way to implement it would be to build a 'target' deceleration profile that gets you to the desired speed in the desired position, and modulate the brakes to achieve that profile, with feedback. That's not too different to what autobrake does, just the target is computer controlled rather than a fixed pre-set rate. Deceleration is obviously never going to perfectly match the target rate so there's going to be some hunting; even autobrake won't be perfectly linear. Achieving 'good enough' is the name of the game in these types of automation problems.
OP's question, to me, is about how the 'target' profile is constructed and whether it's updated continuously or only on touchdown. I have no clue, but I can throw some guesses out there:
- Front-load the deceleration - decelerate early, then taper off as taxi speed is approached. This delivers the most margin for error but leads to the most brake wear and taxi time. It's basically what autobraking until the pilots cancel it delivers.
- Linear deceleration - constant deceleration rate from touchdown to just before turn-off. You'll need to recalculate deceleration rates as the touchdown point moves. Basically just 'selecting the perfect autobrake setting'.
- Back-load the deceleration - maintain a high speed relying mostly on reversers/spoilers until the last moment, then brake hard. This delivers the fastest turn-off time (because you're covering the distance at the highest speed) and puts the least energy into the brakes (because the reversers/spoilers have the most time at high speed). Highest risk of overrun because you're waiting until the last moment to brake.
I expect that #3 most closely matches reality.
There are some technical safeguards that could be added:
- Allow more margin for error (lower assumed deceleration rate or achieve taxi speed with more remaining distance) for full-length landings than intersection turn-offs. It doesn't really matter if you overshoot say 0.1% of landings by 100m/300ft if you're targetting a specific taxiway, but that's a major issue if you're approaching the runway end. IMHO it's a near certainty that this is included.
- Test the brakes/braking action on initial touchdown to check that the runway surface is good enough to achieve the required deceleration rate, then return to the target profile. If the surface isn't good enough (e.g. hydroplaning on what is selected as a dry runway, adjust braking profile to be earlier and lighter, perhaps ask pilots for full reverse.